Dump Data Management for Memory Leak Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for managing dump data in memory leak investigations often result in the deletion of valuable data, making it difficult to analyze the cause of memory leaks due to the inability to determine which data is useful for analysis, leading to storage capacity issues and inefficient use of storage devices.
Innovation Solution
An apparatus that generates object-count information from dump data to identify and preserve dump data with a locally minimum number of objects, excluding it from deletion to ensure that useful data for cause analysis is retained, while managing storage capacity effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dump data is collected frequently to ensure sufficient data for analysis, then the quality of cause analysis is improved, but the storage capacity becomes tight
Solution Approach 1:
The patent extracts only the necessary dump data for analysis by applying deletion rules that identify and remove redundant files. The system extracts useful information (object-count information) from dump data to determine which files should be preserved, thereby reducing storage requirements while maintaining analysis quality.
Solution Approach 2:
The patent changes the parameter of data selection by using object-count information as a criterion for determining which dump data to delete. Instead of deleting based on time or size alone, the system evaluates the object count parameter to identify locally minimum points, which indicate useful data for detecting memory leaks.
2Loss of information
If all collected dump data is stored to preserve useful information, then the completeness of analysis data is improved, but storage utilization becomes inefficient
Solution Approach 1:
The patent introduces feedback mechanisms where object-count information from dump data is used to inform subsequent deletion decisions. The system continuously monitors object counts across multiple dump files and uses this feedback to dynamically determine which files to delete, ensuring that useful information is preserved while improving storage efficiency.
Solution Approach 2:
The system performs self-service by automatically analyzing its own dump data to identify which files are useful and which can be deleted. The dump data management apparatus autonomously evaluates object-count information and applies deletion rules without requiring manual intervention, thereby efficiently managing storage resources.
3Quantity of substance
If dump data is deleted to reduce storage usage, then storage capacity is improved, but valuable analysis data may be lost
Solution Approach 1:
The patent applies preliminary action by generating object-count information from dump data before deletion decisions are made. This preliminary analysis allows the system to identify locally minimum points and determine which files contain useful information for detecting memory leaks, ensuring that valuable data is not lost during the deletion process.
Data Source
AI summary
An apparatus includes a memory to store objects of a plurality of classes, and a storage device to store plural pieces of dump data acquired at different times from the memory. The apparatus generates object-count information indicating a number of objects belonging to each of a plurality of classes, based on the plural pieces of dump data acquired at different times from a memory storing objects of the plurality of classes. The apparatus determines, for each class, at least a portion of dump data that includes a locally minimum number of objects in time variation of the number of objects of the each class, as first dump data that is candidate for storage, based on the generated object-count information. When reducing a total amount of dump data, the apparatus excludes the first dump data determined for each class from second dump data that is to be deleted.